Table 5.
Analysis of interaction ER ratio and ‘over-commitment’ based on the additivity criterion. ELSA-Brasil, 2012–2014.
| Variable | Prev. (%) | PR * | PR ** | 95%CI | Excess of Prevalence e | Measure | 95%CI |
|---|---|---|---|---|---|---|---|
| (1) E-R Ratio = 0, ‘over-commitment’=0 |
3.29 | 1.00 | 1.00 | - | - | ||
| (2) ER Ratio = 1, ‘over-commitment’=0 |
5.65 | 1.71 | 1.71 | 1.29–2.25 | 2.36 | ||
| (3) E-R Ratio = 0, ‘over-commitment’=1 |
5.89 | 1.79 | 2.02 | 1.53–2.68 | 2.60 | ||
| (4) E-R Ratio = 1, ‘over-commitment’ = 1 |
10.79 | 3.28 | 3.68 | 2.97–4.56 | 7.50 | ||
| Expected combined effect a | 8.25 | 2.01 | 2.73 | ||||
| RERI b | 0.945 | 0.81–1.80 | |||||
| AP c | 0.256 | 0.04–0.46 | |||||
| S d | 1.543 | 1.00–2.39 |
Interaction exists if RERI ≠ 0 or AP ≠ 0 or S ≠ 1; p Value for RERI: 0.03; AP: 0.01; and S = 0.05. * PR without adjustment ** Adjusted for age, sex, marital status, race/skin color, education. a Expected combined effect: Prevalence rate = P01 − P00 + P10 − P00 + P00/Prevalence ratio: PR01 − PR00 + PR10 − PR00 + PR00. b Excess risk due to interaction (RERI) = PR11 − PR01 − PR10 + 1. c Attributable proportion due to interaction (AP) = (PR11 − PR01 − PR10 + 1)/PR11. d Synergy index (S) = (PR11 − 1)/(PR01 + PR10 − 2). e Excess of prevalence ratio = (EP = Pexposure − Pno exposure).